Face Generation

In this project, you'll use generative adversarial networks to generate new images of faces.

Get the Data

You'll be using two datasets in this project:

  • MNIST
  • CelebA

Since the celebA dataset is complex and you're doing GANs in a project for the first time, we want you to test your neural network on MNIST before CelebA. Running the GANs on MNIST will allow you to see how well your model trains sooner.

If you're using FloydHub, set data_dir to "/input" and use the FloydHub data ID "R5KrjnANiKVhLWAkpXhNBe".

In [32]:
data_dir = './data'

# FloydHub - Use with data ID "R5KrjnANiKVhLWAkpXhNBe"
# data_dir = '/input'


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import helper

helper.download_extract('mnist', data_dir)
helper.download_extract('celeba', data_dir)
Found mnist Data
Found celeba Data

Explore the Data

MNIST

As you're aware, the MNIST dataset contains images of handwritten digits. You can view the first number of examples by changing show_n_images.

In [33]:
show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
%matplotlib inline
import os
from glob import glob
from matplotlib import pyplot

mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'mnist/*.jpg'))[:show_n_images], 28, 28, 'L')
pyplot.imshow(helper.images_square_grid(mnist_images, 'L'), cmap='gray')
Out[33]:
<matplotlib.image.AxesImage at 0x7fea367fba58>

CelebA

The CelebFaces Attributes Dataset (CelebA) dataset contains over 200,000 celebrity images with annotations. Since you're going to be generating faces, you won't need the annotations. You can view the first number of examples by changing show_n_images.

In [34]:
show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'img_align_celeba/*.jpg'))[:show_n_images], 28, 28, 'RGB')
pyplot.imshow(helper.images_square_grid(mnist_images, 'RGB'))
Out[34]:
<matplotlib.image.AxesImage at 0x7fea35498320>

Preprocess the Data

Since the project's main focus is on building the GANs, we'll preprocess the data for you. The values of the MNIST and CelebA dataset will be in the range of -0.5 to 0.5 of 28x28 dimensional images. The CelebA images will be cropped to remove parts of the image that don't include a face, then resized down to 28x28.

The MNIST images are black and white images with a single color channel while the CelebA images have 3 color channels (RGB color channel).

Build the Neural Network

You'll build the components necessary to build a GANs by implementing the following functions below:

  • model_inputs
  • discriminator
  • generator
  • model_loss
  • model_opt
  • train

Check the Version of TensorFlow and Access to GPU

This will check to make sure you have the correct version of TensorFlow and access to a GPU

In [35]:
"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
from distutils.version import LooseVersion
import warnings
import tensorflow as tf

# Check TensorFlow Version
assert LooseVersion(tf.__version__) >= LooseVersion('1.0'), 'Please use TensorFlow version 1.0 or newer.  You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))

# Check for a GPU
if not tf.test.gpu_device_name():
    warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
    print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))
TensorFlow Version: 1.8.0
Default GPU Device: /device:GPU:0

Input

Implement the model_inputs function to create TF Placeholders for the Neural Network. It should create the following placeholders:

  • Real input images placeholder with rank 4 using image_width, image_height, and image_channels.
  • Z input placeholder with rank 2 using z_dim.
  • Learning rate placeholder with rank 0.

Return the placeholders in the following the tuple (tensor of real input images, tensor of z data)

In [36]:
import problem_unittests as tests

def model_inputs(image_width, image_height, image_channels, z_dim):
    """
    Create the model inputs
    :param image_width: The input image width
    :param image_height: The input image height
    :param image_channels: The number of image channels
    :param z_dim: The dimension of Z
    :return: Tuple of (tensor of real input images, tensor of z data, learning rate)
    """
    # TODO: Implement Function
    inputs_real = tf.placeholder(tf.float32, (None, image_width, image_height, image_channels), name='input_real')
    inputs_z = tf.placeholder(tf.float32, (None, z_dim), name='input_z')
    learning_rate = tf.placeholder(tf.float32, name='learning_rate')
    return inputs_real, inputs_z, learning_rate


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_inputs(model_inputs)
Tests Passed

Discriminator

Implement discriminator to create a discriminator neural network that discriminates on images. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "discriminator" to allow the variables to be reused. The function should return a tuple of (tensor output of the discriminator, tensor logits of the discriminator).

In [37]:
def discriminator(images, reuse=False):
    """
    Create the discriminator network
    :param images: Tensor of input image(s)
    :param reuse: Boolean if the weights should be reused
    :return: Tuple of (tensor output of the discriminator, tensor logits of the discriminator)
    """
    # TODO: Implement Function
    alpha = 0.2
    with tf.variable_scope('discriminator', reuse=reuse):
        x1 = tf.layers.conv2d(images, 64, 5, strides=2, padding='same')
        relu1 = tf.maximum(alpha * x1, x1)
        
        x2 = tf.layers.conv2d(relu1, 128, 5, strides=2, padding='same')
        bn2 = tf.layers.batch_normalization(x2, training=True)
        relu2 = tf.maximum(alpha * bn2, bn2)
        
        x3 = tf.layers.conv2d(relu2, 256, 5, strides=2, padding='same')
        bn3 = tf.layers.batch_normalization(x3, training=True)
        relu3 = tf.maximum(alpha * bn3, bn3)

        # Flatten it
        flat = tf.reshape(relu3, (-1, 4*4*256))
        logits = tf.layers.dense(flat, 1)
        out = tf.sigmoid(logits)
    return out, logits


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_discriminator(discriminator, tf)
Tests Passed

Generator

Implement generator to generate an image using z. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "generator" to allow the variables to be reused. The function should return the generated 28 x 28 x out_channel_dim images.

In [38]:
def generator(z, out_channel_dim, is_train=True):
    """
    Create the generator network
    :param z: Input z
    :param out_channel_dim: The number of channels in the output image
    :param is_train: Boolean if generator is being used for training
    :return: The tensor output of the generator
    """
    # TODO: Implement Function
    alpha = 0.2
    
    # TODO not sure about reuse flag here
    with tf.variable_scope('generator', reuse=(not is_train)):
        # First fully connected layer
        x1 = tf.layers.dense(z, 7 * 7 * 512)
        # Reshape it to start the convolutional stack
        x1 = tf.reshape(x1, (-1, 7, 7, 512))
        x1 = tf.layers.batch_normalization(x1, training=is_train)
        x1 = tf.maximum(alpha * x1, x1)
        
        x2 = tf.layers.conv2d_transpose(x1, 256, 5, strides=2, padding='same')
        x2 = tf.layers.batch_normalization(x2, training=is_train)
        x2 = tf.maximum(alpha * x2, x2)
        
        x3 = tf.layers.conv2d_transpose(x2, 128, 5, strides=2, padding='same')
        x3 = tf.layers.batch_normalization(x3, training=is_train)
        x3 = tf.maximum(alpha * x3, x3)
        
        # Output layer
        logits = tf.layers.conv2d_transpose(x3, out_channel_dim, 4, strides=1, padding='same')
        out = tf.tanh(logits)
    
    return out


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_generator(generator, tf)
Tests Passed

Loss

Implement model_loss to build the GANs for training and calculate the loss. The function should return a tuple of (discriminator loss, generator loss). Use the following functions you implemented:

  • discriminator(images, reuse=False)
  • generator(z, out_channel_dim, is_train=True)
In [39]:
def model_loss(input_real, input_z, out_channel_dim):
    """
    Get the loss for the discriminator and generator
    :param input_real: Images from the real dataset
    :param input_z: Z input
    :param out_channel_dim: The number of channels in the output image
    :return: A tuple of (discriminator loss, generator loss)
    """
    # TODO: Implement Function
    g_model = generator(input_z, out_channel_dim)
    d_model_real, d_logits_real = discriminator(input_real)
    d_model_fake, d_logits_fake = discriminator(g_model, reuse=True)
    
    smooth = 0.1
    
    d_loss_real = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real, labels=tf.ones_like(d_model_real) * (1- smooth)))
    d_loss_fake = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.zeros_like(d_model_fake)))
    g_loss = tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.ones_like(d_model_fake)))

    d_loss = d_loss_real + d_loss_fake

    return d_loss, g_loss


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_loss(model_loss)
Tests Passed

Optimization

Implement model_opt to create the optimization operations for the GANs. Use tf.trainable_variables to get all the trainable variables. Filter the variables with names that are in the discriminator and generator scope names. The function should return a tuple of (discriminator training operation, generator training operation).

In [40]:
def model_opt(d_loss, g_loss, learning_rate, beta1):
    """
    Get optimization operations
    :param d_loss: Discriminator loss Tensor
    :param g_loss: Generator loss Tensor
    :param learning_rate: Learning Rate Placeholder
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :return: A tuple of (discriminator training operation, generator training operation)
    """
    # TODO: Implement Function
    # Get weights and bias to update
    t_vars = tf.trainable_variables()
    d_vars = [var for var in t_vars if var.name.startswith('discriminator')]
    g_vars = [var for var in t_vars if var.name.startswith('generator')]

    # Optimize
    with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)):
        d_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(d_loss, var_list=d_vars)
        g_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(g_loss, var_list=g_vars)

    return d_train_opt, g_train_opt


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_opt(model_opt, tf)
Tests Passed

Neural Network Training

Show Output

Use this function to show the current output of the generator during training. It will help you determine how well the GANs is training.

In [41]:
"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np

def show_generator_output(sess, n_images, input_z, out_channel_dim, image_mode):
    """
    Show example output for the generator
    :param sess: TensorFlow session
    :param n_images: Number of Images to display
    :param input_z: Input Z Tensor
    :param out_channel_dim: The number of channels in the output image
    :param image_mode: The mode to use for images ("RGB" or "L")
    """
    cmap = None if image_mode == 'RGB' else 'gray'
    z_dim = input_z.get_shape().as_list()[-1]
    example_z = np.random.uniform(-1, 1, size=[n_images, z_dim])

    samples = sess.run(
        generator(input_z, out_channel_dim, False),
        feed_dict={input_z: example_z})

    images_grid = helper.images_square_grid(samples, image_mode)
    pyplot.imshow(images_grid, cmap=cmap)
    pyplot.show()

Train

Implement train to build and train the GANs. Use the following functions you implemented:

  • model_inputs(image_width, image_height, image_channels, z_dim)
  • model_loss(input_real, input_z, out_channel_dim)
  • model_opt(d_loss, g_loss, learning_rate, beta1)

Use the show_generator_output to show generator output while you train. Running show_generator_output for every batch will drastically increase training time and increase the size of the notebook. It's recommended to print the generator output every 100 batches.

In [48]:
def scale(x):
    x = ((x - x.min())/(255 - x.min()))
    lower, upper = -1, 1
    return x * (upper - lower) + lower

def train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode):
    """
    Train the GAN
    :param epoch_count: Number of epochs
    :param batch_size: Batch Size
    :param z_dim: Z dimension
    :param learning_rate: Learning Rate
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :param get_batches: Function to get batches
    :param data_shape: Shape of the data
    :param data_image_mode: The image mode to use for images ("RGB" or "L")
    """
    # TODO: Build Model
    
    img_channel_num = 3 if data_image_mode == 'RGB' else 1
    input_real, input_z, lrn_rate = model_inputs(data_shape[1], data_shape[2], img_channel_num, z_dim)
    d_loss, g_loss = model_loss(input_real, input_z, data_shape[3])
    # TODO learning rate VS lrn_rate
    d_opt, g_opt = model_opt(d_loss, g_loss, learning_rate, beta1)
    steps = 0
    show_every = 100
    
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for epoch_i in range(epoch_count):
            for batch_images in get_batches(batch_size):
#                 batch_images = scale(batch_images)
                # TODO: Train Model
                steps += 1
                batch_z = np.random.uniform(-1, 1, size=(batch_size, z_dim))
                
                _ = sess.run(d_opt, feed_dict={input_real: batch_images, input_z: batch_z})
                _ = sess.run(g_opt, feed_dict={input_z: batch_z, input_real: batch_images})
                
                if steps % show_every == 0:
                    # At the end of each epoch, get the losses and print them out
                    train_loss_d = d_loss.eval({input_z: batch_z, input_real: batch_images})
                    train_loss_g = g_loss.eval({input_z: batch_z})

                    print("Discriminator Loss: {:.4f}...".format(train_loss_d),
                          "Generator Loss: {:.4f}".format(train_loss_g))
                    show_generator_output(sess, 16, input_z, img_channel_num, data_image_mode)
                    

MNIST

Test your GANs architecture on MNIST. After 2 epochs, the GANs should be able to generate images that look like handwritten digits. Make sure the loss of the generator is lower than the loss of the discriminator or close to 0.

In [49]:
batch_size = 32
z_dim = 128
learning_rate = 0.0001
beta1 = 0.3


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 2

mnist_dataset = helper.Dataset('mnist', glob(os.path.join(data_dir, 'mnist/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
          mnist_dataset.shape, mnist_dataset.image_mode)
Discriminator Loss: 1.3084... Generator Loss: 1.6344
Discriminator Loss: 0.6169... Generator Loss: 2.7238
Discriminator Loss: 0.5651... Generator Loss: 3.4358
Discriminator Loss: 1.5782... Generator Loss: 0.7068
Discriminator Loss: 1.0070... Generator Loss: 1.1028
Discriminator Loss: 0.6665... Generator Loss: 5.7066
Discriminator Loss: 1.0029... Generator Loss: 4.8244
Discriminator Loss: 0.7053... Generator Loss: 1.9002
Discriminator Loss: 0.9221... Generator Loss: 1.2029
Discriminator Loss: 0.7069... Generator Loss: 2.5987
Discriminator Loss: 0.8622... Generator Loss: 1.3267
Discriminator Loss: 0.5169... Generator Loss: 5.3380
Discriminator Loss: 0.6769... Generator Loss: 2.0935
Discriminator Loss: 0.7432... Generator Loss: 3.4440
Discriminator Loss: 0.7028... Generator Loss: 3.7140
Discriminator Loss: 0.5979... Generator Loss: 2.7412
Discriminator Loss: 0.5541... Generator Loss: 4.6738
Discriminator Loss: 0.5149... Generator Loss: 5.6830
Discriminator Loss: 0.5343... Generator Loss: 4.5583
Discriminator Loss: 0.5375... Generator Loss: 4.2464
Discriminator Loss: 0.5564... Generator Loss: 3.1183
Discriminator Loss: 0.6307... Generator Loss: 2.4270
Discriminator Loss: 0.5193... Generator Loss: 4.6304
Discriminator Loss: 0.9129... Generator Loss: 1.4169
Discriminator Loss: 0.5175... Generator Loss: 4.5509
Discriminator Loss: 0.6145... Generator Loss: 2.4449
Discriminator Loss: 0.5186... Generator Loss: 4.9316
Discriminator Loss: 0.5398... Generator Loss: 3.5930
Discriminator Loss: 0.5488... Generator Loss: 3.5061
Discriminator Loss: 0.5403... Generator Loss: 3.7488
Discriminator Loss: 0.5884... Generator Loss: 2.6824
Discriminator Loss: 0.5423... Generator Loss: 3.8101
Discriminator Loss: 0.9876... Generator Loss: 1.1566
Discriminator Loss: 0.7395... Generator Loss: 1.6722
Discriminator Loss: 0.7019... Generator Loss: 1.8176
Discriminator Loss: 0.5291... Generator Loss: 3.8698
Discriminator Loss: 0.8536... Generator Loss: 1.3453

CelebA

Run your GANs on CelebA. It will take around 20 minutes on the average GPU to run one epoch. You can run the whole epoch or stop when it starts to generate realistic faces.

In [50]:
batch_size = 32
z_dim = 128
learning_rate = 0.002
beta1 = 0.3


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 1

celeba_dataset = helper.Dataset('celeba', glob(os.path.join(data_dir, 'img_align_celeba/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
          celeba_dataset.shape, celeba_dataset.image_mode)
Discriminator Loss: 0.5182... Generator Loss: 4.5408
Discriminator Loss: 0.6141... Generator Loss: 2.4770
Discriminator Loss: 1.2838... Generator Loss: 0.7689
Discriminator Loss: 1.4547... Generator Loss: 0.9662
Discriminator Loss: 1.6349... Generator Loss: 0.4605
Discriminator Loss: 1.2325... Generator Loss: 0.7946
Discriminator Loss: 0.6479... Generator Loss: 4.0866
Discriminator Loss: 1.2166... Generator Loss: 0.9750
Discriminator Loss: 0.8298... Generator Loss: 1.6733
Discriminator Loss: 0.6093... Generator Loss: 2.8202
Discriminator Loss: 2.4118... Generator Loss: 0.1869
Discriminator Loss: 1.2219... Generator Loss: 0.9570
Discriminator Loss: 1.3859... Generator Loss: 0.9106
Discriminator Loss: 1.2740... Generator Loss: 0.7683
Discriminator Loss: 1.5053... Generator Loss: 0.5844
Discriminator Loss: 1.2624... Generator Loss: 1.1744
Discriminator Loss: 1.2813... Generator Loss: 0.8617
Discriminator Loss: 1.2378... Generator Loss: 0.9442
Discriminator Loss: 1.1684... Generator Loss: 1.1590
Discriminator Loss: 1.1995... Generator Loss: 1.2051
Discriminator Loss: 1.3606... Generator Loss: 0.7188
Discriminator Loss: 1.5437... Generator Loss: 0.4942
Discriminator Loss: 1.3818... Generator Loss: 0.7587
Discriminator Loss: 1.3159... Generator Loss: 0.8462
Discriminator Loss: 1.4497... Generator Loss: 0.6162
Discriminator Loss: 1.3426... Generator Loss: 0.8121
Discriminator Loss: 1.3534... Generator Loss: 0.8518
Discriminator Loss: 1.3798... Generator Loss: 0.7983
Discriminator Loss: 1.2564... Generator Loss: 1.0739
Discriminator Loss: 1.3236... Generator Loss: 0.9378
Discriminator Loss: 1.2999... Generator Loss: 0.9876
Discriminator Loss: 1.3328... Generator Loss: 0.9584
Discriminator Loss: 1.3151... Generator Loss: 0.9213
Discriminator Loss: 1.3365... Generator Loss: 0.8282
Discriminator Loss: 1.3849... Generator Loss: 0.7227
Discriminator Loss: 1.4462... Generator Loss: 0.8088
Discriminator Loss: 1.2367... Generator Loss: 0.8844
Discriminator Loss: 1.2996... Generator Loss: 0.8257
Discriminator Loss: 1.4683... Generator Loss: 0.6142
Discriminator Loss: 1.3752... Generator Loss: 0.7288
Discriminator Loss: 1.4096... Generator Loss: 0.6727
Discriminator Loss: 1.4066... Generator Loss: 0.6621
Discriminator Loss: 1.3312... Generator Loss: 0.7988
Discriminator Loss: 1.2872... Generator Loss: 0.9861
Discriminator Loss: 1.3285... Generator Loss: 0.9311
Discriminator Loss: 1.3442... Generator Loss: 0.9131
Discriminator Loss: 1.2968... Generator Loss: 0.8920
Discriminator Loss: 1.3279... Generator Loss: 0.8336
Discriminator Loss: 1.3452... Generator Loss: 0.7330
Discriminator Loss: 1.3618... Generator Loss: 0.8109
Discriminator Loss: 1.3334... Generator Loss: 0.8236
Discriminator Loss: 1.2992... Generator Loss: 0.7884
Discriminator Loss: 1.3204... Generator Loss: 0.8739
Discriminator Loss: 1.2853... Generator Loss: 0.8675
Discriminator Loss: 1.3244... Generator Loss: 0.8253
Discriminator Loss: 1.3343... Generator Loss: 0.7331
Discriminator Loss: 1.3279... Generator Loss: 0.8604
Discriminator Loss: 1.3928... Generator Loss: 0.7551
Discriminator Loss: 1.3738... Generator Loss: 0.7027
Discriminator Loss: 1.2952... Generator Loss: 0.9797
Discriminator Loss: 1.6578... Generator Loss: 1.8607
Discriminator Loss: 1.2863... Generator Loss: 0.8389
Discriminator Loss: 1.2936... Generator Loss: 0.7382

Submitting This Project

When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as "dlnd_face_generation.ipynb" and save it as a HTML file under "File" -> "Download as". Include the "helper.py" and "problem_unittests.py" files in your submission.